Research Methods Hypothesis Test Items

38

3.1. Research Methods

This research employed quantitative analysis with quasi experimental design. The design was chosen to test the hypothesis. This research entangled three classes; the first class was a try-out class; it was used to seek the validity and the reliability of instruments. The second was control class and the third was experimental class Hatsch and Farhady, 1982; Hatsch and Lazaraton, 1994; Kranszler and Moursund, 1999; Dornyei, 2007. The calculation result was analyzed and discussed to answer the research question. Meanwhile, the questionnaire analysis was used to triangulate the data based on the statistical calculation result. The result of questionnaire and vocabulary test were compared and synthesized to get deeper interpretation of the research. In addition, that result was validated using related theoretical foundation of expert views.

3.2. Hypothesis

This research was started with null-hypothesis H where both classes; experimental and control were considered have no significant different in the level of mastering vocabulary. H : µ experimental = µ control 39 In specific, the null hypothesis of this research is playing a computer game cannot develop students’ ability in vocabulary mastery. It means that there is no significant difference between experimental class and control class in mean adjustment level Hatsch and Farhady, 1982; Hatsch and Lazaraton, 1994; Kranszler and Moursund, 1999; Dornyei, 2007. By using null hypothesis, every possibility of the research can be shown. If the hypothesis is rejected, it can be concluded that experiment works. While, if the hypothesis is accepted, the experiment does not work.

3.3. Population and Sample

The main prerequisite of population and sample for this research was the sample should know how to operate computer and engage with them continually. Yet, the population and sample should not have ever played the computer game given before. Because of that, purposive sampling was used. As the result, students of one of junior high school in Sumedang were chosen. The target population for this research is homogeny population. Therefore, the population for this research was the second grades students of junior high school. They were chosen for they become accessible population for this research. So, all second grades students in one of junior high school in Sumedang became the population of this research. 40 Furthermore, the population selected was narrowed to be samples. Samples for this research were taken using purposive sampling. It employed clustered intact group sampling from the population. So that, every class within the population or sample frame has the same chance to be chosen as experimental class and control class. Since there were three classes conducted for this research, two classes were chosen first as experimental class and control class. And the other class performed as try-out class. As a result, 8a was chosen as control class, while 8b took part as experimental class. Both control class and experimental class consist of 20 students. Since the number of samples does not reach the minimum criteria to apply several statistical calculations, two tests were employed; normal distribution test and homogeneity of variance. These tests aim to identify whether t-test calculation continues or not Hatsch and Farhady, 1982; Hatsch and Lazaraton, 1994; Kranszler and Moursund, 1999; Dornyei, 2007.

3.3.1. Try-Out Class

The try-out class was used to analyze the validity and the reliability of the instruments, so that the instruments can be utilized as a measurement to test students’ ability in mastering vocabulary. This class consists of 25 students excluded the control class and the experimental class. The same grade students 41 with control class and experimental class were chosen, in order to eliminate the possible issues which relate to validity and reliability of the instruments.

3.3.2. Experimental Class

The experimental class of this research consists of 20 students who are assumed in the same level in their vocabulary mastery. As it calls, the experiments were given to them in some segments. Generally, a computer game was given to them to be played. Thus, teacher guided them and involved with them to play the game. In order to achieve or to finish the game, teacher did not give the participants any clues about what this game is for. In the process, teacher did not show the way how. Furthermore, teacher did not give students the definition of the unfamiliar words or direct translation. So, students could find a way to understand those words by their effort. The experiment took eight meetings. Hence, the experiment stopped when they finished or completed the game. In addition to those eight meetings, another one meeting was held before the experiment begun, as a pre-test and another one meeting for post-test. In total, there were ten meetings held for completing this research. 42

3.3.3. Control Class

Similar with the experimental class, the control class consists of 20 students. This class performed as a static class that controlled the vocabulary improvement of experimental class. This result of the research was analyzed by comparing the test result with the experimental class. No experiment held in control class. In other words, no changes of way they learned vocabulary. Roughly, they learned vocabulary along with reading comprehension in the class. They read the text and try to guess the meaning of the new words they found in the passage. Otherwise, they wrote down the unknown words, gave the words to their teacher and teacher explained the meaning of the words or direct translation. Then, commonly, teacher asked them to produce sentences or paragraphs consist of those new words. And n the end of week or unit, students were given a quiz or test related to the vocabulary given. Again, it was along with reading the passage or reading the text. As in experimental class this class also held 10 meeting in total. 8 meeting for delivering material to students and the other two were taken for pre-test and post- test. 43

3.4. Research Instruments

This research employed vocabulary test as the main instruments. These instruments were supported by the finding from the questionnaire. Below are the explanations of instruments for this research.

3.4.1. Vocabulary Test

This vocabulary test is a measurement of students’ ability in mastering vocabulary. The test comprises 30 multiple choice question items which were tested to both experimental class and control class. However, to build the test as a good measurement of vocabulary mastering ability, its validity and reliability of the vocabulary tests appeared. In the very beginning, 50 items of multiple choices question were analyzed to seek the highest validity and reliability for every question items. Before they were analyzed, those 50 items of multiple choices question are validated by experts. In consequence, several chances were done to the items. Some of them considered the words which should be familiar and localized based on the syllabus and curriculum of junior high school students in Sumedang. In addition, the site and the condition of students were considered to make the items as valid and reliable as possible. 44

3.4.1.1. Vocabulary Test Item

The test was given to try-out class to seek their validity and reliability of instruments before it was tested to both control class and experimental class. The analysis began with 50 question items of vocabulary test in order to get in order to choose the better validity and reliability scores. All questions items were developed from the school syllabus, relevant theories and part of the conversations taken from the game played. They cover four basic competences and several indicators see Appendix B. Yet, before applying the test to try-out class, the items were analyzed by experts and document analysis by comparing the test item to the syllabus and the specific theories. Therefore, from 11 vocabulary skills which are proposed by Gairns and Redman 1986 cited from Lewis 2001, this test items are covered 10 of them. There are boundaries between conceptual meanings, polysemy, homophony, homonymy, synonymy, affective meaning, style, register and dialect, translation, chunks of language, and grammar of vocabulary. While pronunciation was not included since this skill are related to recognize and to reproduce words in speech. Meanwhile, questions items can be seen in the appendix A. To make clear understanding about the items and prove the validity and reliability of the instruments, the specification used for constructing test can be seen on appendix B. 45

3.4.1.2. Validity test

Arikunto 1993; 63-69 explained that validity is a measurement of instrument. If the validity value of the instrument is lowpoor, the instrument cannot be trusted to measure something. It means the invalid instrument cannot be applied to the respondents. Arikunto also proposed the use of Pearson Product Moment Correlation Coefficient Values to seek the value of instruments’ validity. = ∑ ∑ ∑ ∑ ∑ ∑ ∑ Hatsch and Farhady, 1982; Hatsch and Lazaraton, 1994; Kranszler and Moursund, 1999; Riduwan, 2003; Dornyei, 2007. Where r is Pearson Product Moment Correlation Coefficient Values, N is number of students who are analyzed, x is students’ vocabulary score first variable and y is students’ summative score second variable. Every item on the instrument was calculated by using Pearson Product Moment Correlation Coefficient Values to seek correlation index for every items through correlating every single item of instrument x with total score of instrument y. Next, the r is interpreted as follow : 46 Table 3.1. Index of Validity Level Coefficient Interval r Validity Level 0.800 – 1.000 Excellent Ex 0.600 – 0.799 Good Gd 0.400 – 0.599 Satisfying St 0.200 – 0.399 Poor Pr 0.000 – 0.199 Very Poor Vp Riduwan 2004:110

3.4.1.3. Reliability test

Not only validity analysis but also reliability analyses were employed for this research, so that the effective measurement of vocabulary test can be given to gather data as exact as possible to students’ ability in mastering vocabulary. This research used KUDER RICHARDSON-20 KR-20 method to analyze reliability of instrument. It is because in this research instrument, every single right-answer is valued 1 and every wrong-answer is 0. In addition, this method was used to gain a higher reliability value, since the result of KR-20 tends to give a higher value than the other methods such as KR-21, Anova Hoyt, Alpha and so on. Arikunto 1993: 101. KUDER RICHARDSON-20 KR-20 formula : 47 = − 1 − ∑ r 11 = Internal reliability coefficient for all items K = sum of question item p = proportion of subjects who answer right = + , - q = proportion of subjects who answer wrong q = 1 – p s 2 = variance total = . ∑ − ∑ . . − 1 Afterward, the value of r 11 is compared with index of Pearson Product Moment Correlation Coefficient Values see Appendix E to see whether the value is reliable or not. 48

3.4.1.4. The Result of the Try-Out test

On Monday, May 2, 2011, the instrument was applied to try out class, eight grade students of one of junior secondary school in Sumedang, to seek the validity and reliability of the instrument. The result is shown as follow.

a. Test Items

The calculation showed that from 50 questions items of vocabulary tested, 34 items were categorized valid and 16 items were invalid. So that 30 valid items of vocabulary test were taken as the instrument of this research and the instrument was covered as shown on the table below. Table 3.2. Index of Validity for Question Items Try-Out test Coefficient Interval r Validity Level Question Item number 0.800 – 1.000 Excellent Ex - 0.600 – 0.799 Good Gd 16,20,27, 0.400 – 0.599 Satisfying St 2,3,5,6,9,10,15,17,18,19,22, 25,26,31,32,33,36,37,38,39, 40,42,46,47,48. 0.200 – 0.399 Poor Pr 11,21,24,28. 0.000 – 0.199 Very Poor Vp 1,13. 0.00 – 0.000 Invalid 4,7,8,12,14,23,29,30,34,35, 41,43,44,45,49,50. 49 Later, the invalid items were eliminated, so that the vocabulary test item consist of three items categorized good, 25 items categorized satisfying and two items categorized poor. Furthermore, after comprising the items into 30 see appendix I, below is the distribution table for each items of pre-test and post-test. Table 3.3. Index of Validity for Question Items Pre-and Post- test Coefficient Interval r Validity Level Question Item number 0.800 – 1.000 Excellent Ex - 0.600 – 0.799 Good Gd 8,12,17 0.400 – 0.599 Satisfying St 1,2,3,4,5,6,7,9,10,11,14,15, 16,19,20,21,22,23,24,25,26, 27,28,29,30 0.200 – 0.399 Poor Pr 13,18 0.000 – 0.199 Very Poor Vp - 0.00 – 0.000 Invalid - In consequence, the distribution of standard competence, basic competence, learning activities and also vocabulary skill offered by the question items were changing as in appendix I.

b. Reliability of the Instrument